基于最大似然估计算法的子孔径拼接检测技术  

Sub-aperture stitching testing technology based on maximum likelihood estimation algorithm

作　　者：闫公敬[1] 张宪忠[1] 

机构地区：[1]齐齐哈尔大学应用技术学院,黑龙江齐齐哈尔161000

出　　处：《红外与激光工程》2015年第7期2206-2210,共5页Infrared and Laser Engineering

基　　金：国家自然科学基金(61036015);国家高技术研究发展计划(O8663NJ090)

摘　　要：大口径平面镜作为光学系统的重要组成部分,其面形精度对系统成像具有重要影响。子孔径拼接检测作为大口径光学平面反射镜检测的常用手段,子孔径拼接算法是该技术的核心。研究了平面子孔径拼接算法,基于最大似然估计与正交化Zernike多项式拟合建立了一套合理的拼接算法与数学模型,基于该算法模型可以有效实现对大口径平面镜的拼接检测,同时编写了相应的拼接程序,并利用Φ100 mm干涉仪对Φ120 mm的平面镜进行了拼接检测,给出了拼接检测与全口径检测的对比结果,对比结果表明:拼接所得全孔径相位分布与全口径检测结果的RMS值偏差分别为0.002λ,验证了算法的可靠性与准确性。As an important part of the optical system, the accuracy of the plane mirror is an influence factor to system imaging. Subaperture stitching testing is a usual way to test plane mirror in large aperture, while the stitching algorithm is the key in the stitching technology. The plane sub-aperture stitching algorithm was studied in the paper and a reasonable stitching algorithms and mathematical models was established based on maximum likelihood estimation and orthogonalization Zernike polynomial fitting. Stitching to plane mirror in large aperture can be accomplished with the above stitching model.Stitching program was also written and stitching testing was carried on with a Φ100 mm interferometer on a Φ120 mm plane mirror. Comparing the stitching result with the full aperture testing result, it shows that the stitching map is in consistent with the full aperture testing map. The difference of RMS between them is 0.002λ, verifying the reliability and accuracy of the algorithm.

关 键 词：干涉检测 子孔径拼接 最大似然估计 Zernike多项式拟合 正交化 

分 类 号：O439[机械工程—光学工程] O436.1[理学—光学]